PacificNorth America plate motion and opening of the Upper Delfn
basin, northern Gulf of California, Mexico
Michael Oskin
Institute for Crustal Studies, University of California, Santa Barbara, California 93106-1100, USA
Joann Stock
California Institute of Technology, Pasadena, California 91125, USA
ABSTRACT
Correlation of conjugate rifted margins
of the Upper Delfn basin constrains the
timing and amount of transtensional open-
ing along the PacificNorth America plate
boundary in the northern Gulf of Califor-
nia. Lithologic, geochemical, paleomagnet-
ic, and geochronologic data from a set of
four ignimbrites, consisting of eight distinc-
tive cooling units, are shown to correlate
from northeastern Baja California to Isla
Tiburon and adjacent areas of western So-
nora. These matching ignimbrites are the
ca. 12.6 Ma tuff of San Felipe, the 6.3 6 0.2
Ma tuffs of Mesa Cuadrada (Tmr3 and
Tmr4), the tuffs of Dead Battery Canyon
(Tmr5), and the 6.1 6 0.5 Ma tuffs of Ar-
royo El Canelo. Offset distributions and fa-
cies patterns of these ignimbrites support
255 6 10 km of opening between conjugate
rifted margins of the Upper Delfn basin.
Addition of deformation from the continen-
tal margins of this basin indicates at least
276 6 13 km of PacificNorth America
plate motion between coastal Sonora and
the main gulf escarpment in Baja Califor-
nia since ca. 6 Ma; a further 20 6 10 km
of northwestward displacement of Isla Ti-
buron relative to coastal Sonora occurred
sometime after 12.6 Ma. These reconstruc-
tions agree with earlier estimates of slip
across the Gulf of California and on the
San Andreas fault system of southern Cal-
ifornia, but require that the PacificNorth
America plate boundary became localized
in the gulf at ca. 6 Ma. The restored con-
tinental margins of the Upper Delfn basin
show that only a 2025 km width of upper
continental crust has foundered beneath
this part of the northern Gulf of California.
E-mail: oskin@crustal.ucsb.edu.
GSA Bulletin; October 2003; v. 115; no. 10; p. 11731190; 12 figures; 3 tables; Data Repository item 2003134.
This result suggests that most of the crustal
area formed by opening of the Upper Delfn
basin was either exhumed from lower-
crustal levels or is new transitional oceanic
crust.
Keywords: Gulf of California, Mexico, rift-
ing, ignimbrite, plate motion, palinspastic
restoration.
INTRODUCTION
The Gulf of California of southwestern
North America (Fig. 1) presents an exception-
al opportunity to examine the formation of a
young ocean basin within an evolving conti-
nental orogen. Opening of the Gulf of Cali-
fornia is the result of transform motion and
extension between the Pacific plate and the
North America plate (Atwater, 1970). The tec-
tonic setting of the Gulf of California and
southwestern North America (Fig. 1) is well
understood from the perspective of continental
geology (Burchfiel et al., 1992; Stewart, 1998;
Stock and Hodges, 1989; Gastil et al., 1991)
and from marine plate-tectonic studies (Stock
and Molnar, 1988; Lonsdale, 1989; Atwater
and Stock, 1998). Opening of the Gulf of Cal-
ifornia followed the termination of subduction
west of Baja California Sur at 12.5 Ma (Mam-
merickx and Klitgord, 1982; Spencer and Nor-
mark, 1979). Magnetic lineations at the mouth
of the Gulf of California record .85% of
PacificNorth America plate motion since 3.5
Ma and .95% of the plate motion since 1 Ma
(DeMets, 1995; DeMets and Dixon, 1999;
Dixon et al., 2000).
Continental geologic records define the tec-
tonic history of the Gulf of California prior to
3.5 Ma. Offset Paleozoic metasedimentary
rocks, Mesozoic batholithic rocks, and Tertia-
ry conglomerates indicate that ;300 km of
For permission to copy, contact editing@geosociety.org
q 2003 Geological Society of America
dextral displacement has occurred in the Gulf
of California (Silver and Chappell, 1988; Gas-
til et al., 1991, 1973) (Fig. 1). Geologic rela-
tionships in southern California indicate that
most or all of this displacement occurred after
12 Ma (Ehlig et al., 1975; Crowell, 1981).
Correlation of synrift pyroclastic deposits
from the Puertecitos volcanic province of Baja
California to Isla Tiburon and coastal Sonora
supports 255 6 10 km of opening of the Up-
per Delfn basin of the northern Gulf of Cal-
ifornia (Oskin et al., 2001) (Fig. 2). These re-
sults suggest that the PacificNorth America
plate boundary became localized in the Gulf
of California during latest Miocene time. With
adequate geologic control, fundamental tec-
tonic problems such as the roles of strain rate,
preexisting weaknesses, and strain partitioning
in localizing continental rifting may be ad-
dressed from the geology of the Gulf of Cal-
ifornia. Critical geologic parameters from the
Gulf of California are the rate of transfer of
Baja California to the Pacific plate and the
amount of crustal attenuation that accompa-
nied this transfer. Until recently, no record has
been documented that directly measured total
offset, offset rate, and crustal attenuation in
the Gulf of California.
This paper expands on the initial findings
reported by Oskin et al. (2001) to document a
complete record of PacificNorth America
plate motion in the northern Gulf of California
and its surrounding rifted continental margin.
Four principal pyroclastic flow deposits, con-
sisting of a total of eight separate cooling
units, are correlated from Baja California to
Isla Tiburon and coastal Sonora. Oskin et al.
(2001) described initial results from matching
two of these pyroclastic flow deposits: the ca.
12.6 Ma tuff of San Felipe, consisting of a
single cooling unit (Stock et al., 1999), and
the 6.3 6 0.2 Ma tuffs of Mesa Cuadrada
1173
 OSKIN and STOCK

and structure of continental rifting in the

northern Gulf of California.

GEOLOGIC MAPPING AND

CORRELATION

Previous Studies

The geology of conjugate rifted margins of

the northern Gulf of California presents sev-
eral possibilities for precise temporal con-
straints on the development of the gulf rift.
Pre- and synrift volcanic deposits of the Puer-
tecitos volcanic province of northern Baja
California comprise a well-dated middle Mio-
cene to late Pliocene stratigraphy adjacent to
the Gulf of California (Stock, 1989; Stock et
al., 1991; Martn-Barajas and Stock, 1993;
Martn-Barajas et al., 1995; Lewis, 1996;
Nagy et al., 1999; Oskin and Stock, 2003b).
Contained within these strata are extensive ig-
nimbrite sequences deposited at 12.6 Ma, ca.
6 Ma, and ca. 3 Ma (Stock et al., 1999; Stock,
2000). Detailed mapping in the northern Puer-
tecitos volcanic province indicates sources for
these deposits near the present-day western
shore of the Gulf of California (Stock et al.,
1991; Martn-Barajas et al., 1995; Lewis,
1994; Nagy, 1997). These widespread ignim-
brites in Baja California with eastern sources
comprise ideal candidates for cross-gulf cor-
relation (Stock et al., 1999; Nagy et al., 1999).
Figure 1. Tectonic map of southwestern North America. Shaded areas represent Basin A likely target for correlation is the coastal
and Range province (light gray) and Gulf extensional province (dark gray). Present-day area of central Sonora, located ;300 km
plate boundaries shown as dark lines. Inactive plate boundaries shown as gray lines. southeast of the Puertecitos area. Reconnais-
Distinctive conglomerate outcrops denoted by circled letters: PPoway conglomerate of sance mapping of coastal Sonora indicates a
Abbott and Smith (1989); FConglomerate with distinctive Permian fusulinid-bearing similar, extensive middle to late Miocene vol-
limestone clasts of Gastil et al. (1973). ABAgua Blanca fault; EElsinore fault; ETR canic cover adjacent to the eastern shoreline
eastern Transverse Ranges; SGSan Gabriel fault; SJSan Jacinto fault; STSalton of the Gulf of California (Gastil and Krum-
Trough; WTRwestern Transverse Ranges. menacher, 1974, 1977).
Remapping of a distinctive Tertiary con-
glomeratefirst correlated across the Gulf of
(new name), consisting of two primary cool- posed earlier by Oskin et al. (2001). Defor- California by Gastil et al. (1973)supports
ing units named Tmr3 and Tmr4 (Stock, 1989; mation of the Sonora continental margin in- that this unit is a robust geologic tie point.
Lewis, 1996; Nagy et al., 1999). The study dicates that an additional 20 6 10 km of Inspection of the basement to cover transition
presented here elaborates on these correlations dextral displacement occurred here sometime over a large area of both northeastern Baja
and establishes additional matches of one after 12.6 Ma. Together with earlier studies of California and coastal Sonora confirms that
cooling unit of the ca. 6.3 Ma tuffs of Dead rotational deformation of the continental mar- outcrops of the distinctive conglomerate are
Battery Canyon (Lewis, 1996) and four cool- gin of Baja California by Lewis and Stock limited to the two previously mapped areas
ing units of the 6.1 6 0.5 Ma tuffs of Arroyo (1998a), the total opening measured across the (Gastil and Krummenacher, 1974; Bryant,
El Canelo (Stock et al., 1991; Martn-Barajas northern Gulf of California of 296 6 17 km 1986; Oskin and Stock, 2003b; distinctive
and Stock, 1993; Nagy et al., 1999). is indistinguishable from prelate Miocene conglomerate in Fig. 2). Outcrop patterns in
Each of the pyroclastic flow deposits cor- geologic tie points. Restoration of the conju- the Sierra Seri of coastal Sonora suggest a sin-
related across the northern Gulf of California gate rifted margins of the Upper Delfn basin gle southwest-directed channel. Outcrops in
displays a unique matching distribution in also constrains the amount of foundered upper Baja California appear to have been isolated
northeastern Baja California and Isla Tiburon. continental crust that may underlie this seg- as a terrace deposit by 12.6 Ma. Correlation
Offset distributions and facies patterns of ment of the Gulf of California. This slip his- of these outcrops suggests ;300 km of dis-
these ignimbrites support 255 6 10 km of tory and crustal budget comprise important placement across the Gulf of California (Gas-
opening of the Upper Delfn basin of the constraints for the tectonic evolution of south- til et al., 1973). The uncertainty of this dis-
northern Gulf of California (Fig. 2), as pro- western North America and the mechanisms placement is difficult to estimate because only

1174 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

The Tuff of San Felipe

The tuff of San Felipe (Figs. 3 and 4) may

be recognized in the field from its distinctive
welding zonation, phenocrysts, and rhyolite
lava inclusions (Table 1). Geochemical anal-
yses consistently show distinctive anorthoclase
and zoned pyroxene phenocrysts (Table 1) and
high Nb, Ce, Rb, and La (Table 2). Pri-
mary thermal remanent magnetization of the
tuff of San Felipe preserves an unusual low-
inclination reversed-polarity direction (Stock
et al., 1999; Oskin et al., 2001). Most sample
localities of the tuff of San Felipe in Baja Cal-
ifornia and Sonora yielded consistent paleo-
magnetic results with evidence for clockwise
vertical-axis rotation of as much as 508 be-
tween sample localities (Lewis and Stock,
1998a; Nagy, 2000; Table DR1).1
Isotopic ages of the tuff of San Felipe and
other bracketing units in Baja California are
Figure 2. Index map of the northern Gulf of California. Marine faults from Fenby and
consistent with an age of ca. 12.6 Ma for this
Gastil (1991). Areas of reconnaissance mapping for this study shown by large dashed
tuff, although individual, well-dated samples
boxes. Areas of detailed mapping shown by smaller dashed lines. Boundaries of the Gulf
range from 13.0 6 0.2 Ma to 10.6 6 0.1 Ma
extensional province shown by locations of major bounding normal faults. Darker gray
(Stock et al., 1999). Ages spanning this range
area is the Puertecitos volcanic province (PVP). Detailed mapping locations: SRBSanta
but with high uncertainty (9.7 6 1.3 Ma to
Rosa Basin (Bryant, 1986); SFSierra San Fermn (Lewis, 1994); VCValle Chico
13.9 6 2.2 Ma) have been measured from the
(Stock, 1993); MTArroyo Matom Transect (Stock et al., 1991, and unpublished map-
tuff of San Felipe in coastal Sonora and Isla
ping); SISanta Isabel Wash (Nagy, 1997); PPuertecitos (Martn-Barajas and Stock,
Tiburon (Oskin, 2002). Argon loss may have
1993); MAMesa El Avion (Stock, unpublished mapping); CICinco Islas (Oskin, 2002);
contributed to anomalously younger ages.
CScoastal Sonora (Oskin, 2002); ITIsla Tiburon (Oskin, 2002). Distinctive conglom-
Older ages are associated with thick, higher-
erate outcrops of Gastil et al. (1973) shown as black squares (SRB and SS). Towns shown
grade tuff outcrops with a higher percentage
as white circles. BKBaha Kino; PCPunta Chueca; SSSierra Seri. F.Z.Fracture
of rhyolite inclusions. Macroscopic evidence
Zone; T.F.transform fault.
for plastic deformation and petrographic evi-
dence for disaggregation of these inclusions
remnant outcrops exist in Baja California and posits to match with the conjugate rifted mar- suggest that they may be contaminating the
the transport distance and path between the gin in Baja California. Interior areas of the age data, as first proposed by Lewis (1994).
correlative outcrops is unknown. The age of southern and northeastern parts of the island Additional study is under way to separately
these outcrops is known only to postdate the were explored in order to assess important date these inclusions and to clarify the isoto-
underlying Mesozoic batholithic rocks and stratigraphic features identified by initial re- pic age of the tuff of San Felipe.
predate a 15 Ma basalt flow (Bryant, 1986). connaissance studies. The interior of the is- The tuff of San Felipe provides a robust
The ;300 km offset of these strata therefore land was mapped primarily by aerial photo- match across the northern Gulf of California.
approximates the total dextral displacement graph and Landsat image interpretation. At 160 km3, it forms the largest-volume py-
across the Gulf of California since at least Stratigraphic and structural complexities ap- roclastic flow in the study area and covers
middle Miocene time. .4000 km2 (Fig. 4 and Table 3). This ignim-
pear most commonly near the western coast-
brite is likely to have erupted from near the
line of Isla Tiburon, whereas much of the in-
Geologic Mapping of Isla Tiburon and present-day mainland Sonora coastline at Pun-
terior area consists of structurally simple,
Coastal Sonora ta Chueca (Fig. 4). The ca. 12.6 Ma age and
well-exposed block-faulted ridges.
stratigraphic position of the tuff of San Felipe
Preliminary correlations of ash-flow tuffs make this ignimbrite ideal to measure the total
Accessible outcrops of ash-flow tuffs
mapped by Gastil and Krummenacher (1974) were established in the field by lithology and dextral displacement across the Gulf of Cali-
were assessed for lithologic similarity to the verified by petrographic examination of phe- fornia since the onset of rifting after 12 to 11
distinctive tuffs mapped in northeastern Baja nocryst, glass, pumice, and lithic content. Fur- Ma (Stock et al., 1999).
California. Potentially correlative tuff out- ther geochemical and paleomagnetic tests Distribution of the tuff of San Felipe west
crops were located on western and northern were applied to the most widespread and strat- of coastal Sonora occurred through west-
Isla Tiburon and adjacent areas of coastal So- igraphically simple ignimbrites common to
nora. Detailed geologic mapping of Isla Ti- northeastern Baja California and western So- 1
GSA Data Repository item 2003134, tuff paleo-
nora (Oskin et al., 2001). Important character- magnetic remanence values and declination anom-
buron concentrated on the western coast of the alies, is available on the Web at http://www.
island. This area presented a well-exposed istics of correlative pyroclastic flow deposits geosociety.org/pubs/ft2003.htm. Requests may also
north-south transect of pyroclastic flow de- are summarized in Tables 1 and 2. be sent to editing@geosociety.org.

Geological Society of America Bulletin, October 2003 1175

 OSKIN and STOCK

TABLE 1. CHARACTERISTICS OF CORRELATIVE PYROCLASTIC FLOW DEPOSITS FROM CONJUGATE RIFTED MARGINS OF THE UPPER DELFIN BASIN

Units Correlative units Lithology, phenocryst chemistry, and paleomagnetism Age

Tuffs of Arroyo El Canelo
Flagpole member Flagpole Tuff (Nagy et al., Thin, indurated, densely welded dark purple to dark orange crystal-rich tuff. Large vapor- No data
1999) phasereplaced fiamme and centimeter-scale variation in plagioclase phenocryst
abundance similar to El Canelo member. Trace amounts of olivine (fayalite?).
Bighorn Sheep Bighorn Sheep Tuff (Nagy et Weakly indurated, crystal and lithic rich. 2%50% 23 cm lithic fragments. Elongate dark No data
member al., 1999) pumice or scoria up to 50 cm in length comprises up to 100% of the unit locally. Matrix,
where present, of pale pink to purple or yellow devitrified ash and 5%10% phenocrysts
(plagioclase 1 anorthoclase(?) . iron-titanium oxides . altered mafic silicates
(pyroxene?)).
El Canelo Tuff of El Canelo (Nagy et al., Densely welded lithic-phyric ash-flow tuff, multiple cooling units, 3%10% phenocrysts 6.1 6 0.5 Ma
member(s) 1999); Tmc2 and Tmc3 (plagioclase 1 anorthoclase .. orthopyroxene iron-titanium oxides . hornblende), dark (Nagy et al.,
(Martn-Barajas et al., purple matrix of welded glass, distinctive pumice fiamme up to 50 cm in length. Basal dark 1999)
1995); t9 and t12 (Stock et purple indurated vitrophyre with centimeter-scale variations in phenocryst concentration and
al., 1991); Tmr8(?) (Lewis, black fiamme. Above basal zones, phenocryst concentration levels out at ;10%, and
1996) pumice vapor-phase altered. Thick high-grade welded sections display extremely thin (,1
cm) fiamme with rheomorphic flow lineations. Internal cooling breaks distinguished by a
volcanic lithic concentration zone overlain by thin, dark, nonwelded tuff and/or dark purple
indurated vitrophyre.
El Oculto member Tuff of Arroyo El Oculto (Nagy Nonwelded lithic-rich devitrified ash-flow tuff, 2% phenocrysts (plagioclase 1 anorthoclase(?) No data
et al., 1999); t14 (Stock et . orthopyroxene), typically 02 m thick, distinctive bright-white weathered surface. Thicker
al., 1991) sections described by Nagy (1997) display a densely welded interior zone, up to decimeter-
sized lithic fragments, and lithic-concentration zones.
Tuffs of Dead Battery Canyon
Tmr6 member Probably not present on Isla 10% flattened pumice, 10%15% phenocrysts (feldspar .. iron-titanium oxides . No data
Tiburon clinopyroxene basaltic hornblende). Indurated dark reddish-brown porcelaneous welded
base capped by lighter-colored vapor-phase altered zones. Normal-polarity, moderate-
inclination paleomagnetic remanence, steeper and clockwise of Tmr5 (Lewis and Stock,
1998a; Table DR1).
Tmr5 member Tmr4a (Lewis, 1996) 3% centimeter-sized pumice, 5%10% phenocrysts (feldspar .. iron-titanium oxides . No data
clinopyroxene). Indurated, dark reddish-brown porcelaneous welded base capped by light
purple vapor-phasealtered tuff. Normal-polarity, moderate-inclination paleomagnetic
remanence (Lewis and Stock, 1998a) (Table DR1 [see footnote 1 in text]).
Tuffs of Mesa Cuadrada
Tmr4 member Densely welded ash-flow tuff, distinctive welding zonation, 1%3% phenocrysts (plagioclase No data
. pyroxene . quartz), 1% volcanic lithic fragments, even-textured porcelaneous matrix.
Thin (,1 m) basal yellow ash and pumice grades into brown to black vitrophyre with
pumice fiamme. Sharp break separates vitrophyre from densely welded red tuff, capped by
vapor-phase altered light purple welded tuff. Thicker sections capped by an additional zone
of nonwelded tuff. Moderate-inclination normal-polarity remanent magnetization, shallower
inclination, and counterclockwise declination relative to Tmr3 (Lewis and Stock, 1998a;
Nagy, 2000; Oskin et al., 2001).
Tm3t member Up to 200 m (typically 02 m) of primary and reworked air-fall, volcaniclastic debris-flow, and No data
pyroclastic flow deposits.
Tmr3 member Tmr3b (Lewis, 1996) Tmr3 Nonwelded to partially welded, 5%10% centimeter-sized volcanic lithic fragments, 10%15% ca. 6.3 6 0.2 Ma,
(type 2) (Nagy et al., 1999) phenocrysts (alkali feldspar .. augite quartz . biotite iron-titanium oxides . basaltic (Nagy et al.,
hornblende fayalite zircon), up to 10% pumice lapilli. Typical deposits are light grayish 1999); 6.2 6 1.0
purple, weathering orange. More welded tuff is bright orange to red, with a dark brown and 5.3 6 2.2
indurated vitrophyre near base. Sanidine and anorthoclase, Ab60An3Or37 to Ab72An13Or15, 0.2 Ma (Oskin,
wt% Fe and 0.3 wt% Ba. Pyroxene, 90% unzoned augite, En7Fs48Wo45, 0.4% Na, 0.3% Ti, 2002); 5.7 6 0.6
0.3% Al, 0.7% Mn. Cores of zoned pyroxenes are Fe-rich enstatite, En57Fs40Wo3, to Ma (Gastil et
ferrosilite, En20Fs76Wo4, ,0.1 wt% Na, ;1.3 wt% Mn, surrounded by Fe-rich augite, En25 to al.,1979)
En7. Moderate-inclination normal-polarity remanent magnetization (Lewis and Stock, 1998a;
Nagy, 2000; Oskin et al., 2001).
Tuff of San Felipe See Stock et al. (1999) for Densely welded, 10%15% phenocrysts (anorthoclase .. augite iron-titanium oxides), Ca. 12.6 Ma
correlation of units in Baja rare accidental lithic fragments, distinctive, rare to uncommon dark rhyolite lava inclusions. (Stock et al.,
California Inclusions contain 20% phenocrysts (alkali feldspar .. phenocrysts . fayalite), form pods 1999); 9.7 6 1.3
up to 40 cm in length, are abundant in thick, high-grade welded-tuff outcrops. Typical Ma and 13.9 6
section develops dark brown to black basal vitrophyre, grading into densely welded, red to 2.2 Ma (Oskin,
orange eutaxitically foliated tuff. Thicker deposits grade into densely welded, red to light 2002)
purple rheomorphic tuff with pumice fiamme up to 50 cm long. Upper part of tuff generally
light-colored and vapor-phase recrystallized. Nonwelded air-fall ash up to 1 m thick may be
present beneath thicker sections. Cryptoperthitic anorthoclase, Ab52An5Or43 to Ab67An15Or18,
0.3 wt% Fe, Ba may be present just above 0.1 wt% detection limit. Pyroxene, 80%90%
unzoned green Fe-rich augite, En7Fs48Wo45, 0.3% Na, 0.2% Ti, 0.4% Al, and 1.0% Mn.
Cores of zoned pyroxenes are brown-colored magnesian augite (up to En30) with higher Al,
lower Na, and variable Mn. Low-inclination, reversed-polarity remanent magnetism differs
substantially from the expected Miocene field direction and probably records an excursion
or transitional field (Stock et al., 1999; Oskin et al., 2001).

trending paleocanyons on Isla Tiburon and in
the northern Puertecitos volcanic province. A
single 3-km-wide paleocanyon in the northern
Sierra Kunkaak is the only known pathway for
the tuff of San Felipe between coastal Sonora
and western Isla Tiburon (Fig. 4). Additional
pathways may exist in unmapped areas of the
Sierra Kunkaak or beneath the channel be-
tween Isla Tiburon and the mainland. On
western Isla Tiburon, the tuff of San Felipe
fills west-trending channels cut into volcanic
and volcaniclastic strata in the south and base-
ment in the north. On the southern part of the
conjugate margin in northeastern Baja Cali-
fornia, the tuff of San Felipe also occupies

1176 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

TABLE 2. TRACE ELEMENT ANALYSES AND COMPARISON OF CORRELATIVE TUFFS Sierra San Fermn instead of a range-wide av-
Trace Tuff of San Felipe Tmr3 Tmr4 Other erage, we find that the inclination and decli-
element PVP nation of Tmr3 (type 2) and Tmr3b relative to
(ppm) Coastal Isla Baja S. San Isla S. San Isla rhyolites##
Sonora Tiburon California Fermn# Tiburon Fermn Tiburon Tmr4 are consistent throughout the Baja Cal-
Y 51.8 52.6 48.151.9 53.6 55.4 26.7 27.6 2057 ifornia margin and consistent with Tmr3 out-
U 5 5 45 4 4 3 4 04 crops on Isla Tiburon (Oskin et al., 2001).
Rb 176.3 195.4 166188 155.6 149.8 124.9 129.9 80127
Th 17 18 1618 12 12 10 10 413
Distinct remanent paleomagnetic directions
Pb 28 25 2728 29 19 23 15 828 from unit Tmr3 (type 1) of Santa Isabel Wash
Ga 21 22 020 23 23 16 16 1621 and unit Tmr3a of the Sierra San Fermn sup-
Nb 24.6 25 22.524.4 20.5 21.1 11.9 12.6 7.619.5
Zr 315 321 384402 329 320 168 177 98432 port the interpretation that these units are in-
Sr 90 35 1960 65 44 88 68 72263 dividual, local members of the tuffs of Mesa
Zn 91 85 5776 107 109 68 42 2480 Cuadrada.
Ni 2 1 89 1 2 1 1 316
Cr 2 6 02 0 4 2 6 08 The distribution and facies of the tuffs of
V 2 2 758 4 2 3 3 042 Mesa Cuadrada correlate closely between
Ce 123 127 111120 94 94 56 57 2087
Ba 78 39 5256 428 460 1093 996 7051184
northeastern Baja California and Isla Tiburon.
La 54 56 5456 43 38 27 26 1344 These ignimbrites comprise the largest of the
Note: Trace element analyses (XRF [X-ray fluorescence]) were done at the University of Massachusetts, ca. 6.3 to 6.1 Ma tuffs in both volume and areal
Amherst. See Rhodes (1988) for typical values of analytical precision from this facility. Analyses reprinted from extent, covering over 2100 km2 with ;120 km3
Oskin et al. (2001) Data Repository. of pyroclastic deposits (Table 3). The northern

Sample number BK-99-05, 28.8808N, 112.0018W.

Sample number BV-99-30, 28.9798N, 112.4588W. extent of outcrops is well exposed in the north-

Range of values for samples of the Tuff of San Felipe from Baja California (Stock et al., 1999). ern Sierra San Fermn and Sierra San Felipe of
#
Sample number SF-92-64, 30.5948N, 114.7568W.

Sample number BV-99-08, 28.9178N, 112.4478W.
Baja California and the Sierra Alta of Isla Ti-

Sample number SF-92-101, 30.5668N, 114.7908W. buron (Fig. 6). The southern extent of outcrops

Sample number TIB-98-17, 28.9798N, 112.4768W. is well exposed in the Sierra Menor of Isla Ti-
##
Range of values for other high-SiO2 rhyolite samples from the Puertecitos area reported by Martn-Barajas
et al. (1995). PVPPuertecitos volcanic province. buron but obscured by younger volcanism in
the Puertecitos area of Baja California. Thick,
higher-grade deposits of the tuffs of Mesa
Cuadrada crop out adjacent to the eastern part
paleocanyons cut into volcaniclastic strata citos volcanic province measured by Martn- of Arroyo Matom, in the southern Sierra San
(Fig. 4). In the northern Sierra San Fermn and Barajas et al. (1995). The concentrations of Fermn, and in the Santa Isabel Wash area in
the adjacent Sierra San Felipe, the tuff of San these and other trace elements in Tmr3 are Baja California and the central Sierra Menor of
Felipe formed a sheet ponded over basement distinct from the tuff of San Felipe, as are the Isla Tiburon (Fig. 6).
rock, arkosic sandstone, and locally derived compositions of feldspar and pyroxene phe- Facies changes within the tuffs of Mesa
fluvial deposits. The thickest and highest- nocrysts (Oskin et al., 2001). Isotopic ages of Cuadrada associated with younger rhyolitic
grade deposits of the tuff of San Felipe occur Tmr3 from the Isla Tiburon overlap isotopic volcanism suggest a vent for this ignimbrite
in the northern Sierra San Fermn and adjacent ages of 6.3 6 0.2 Ma for Tmr3 from Baja located within a 2030-km-diameter area cen-
parts of the Sierra San Felipe, northern Isla California (Stock, 1989; Lewis, 1996; Nagy et tered at the eastern end of Arroyo Matom
Tiburon, and mainland coastal Sonora (Fig. 4). al., 1999; Oskin, 2002) (Table 1). A sample (Fig. 6). In Baja California, abrupt thickness
The southern limit of the tuff of San Felipe of Tmr3 from western Isla Tiburon (Fig. 6) changes in higher-grade Tmr4 occur southeast
may also correlate across the Gulf of Califor- was dated by Gastil et al. (1979) at 5.7 6 0.6 of Mesa Cuadrada, at Santa Isabel Wash, and
nia, but a cover of younger volcanic rocks Ma (sample no. 1012, K-Ar on feldspar). in the southern Sierra San Fermn. These
limits exposures. The northern limit of the tuff On the basis of paleomagnetic measure- abrupt thickness changes may reflect under-
of San Felipe on the Sonoran margin appears ments, Nagy (2000) distinguished two types lying topographic relief caused by subsidence
to have been controlled by paleotopography. of Tmr3 and suggested that neither Tmr3 due to a prior eruption (e.g., Tmr3) and are
Outcrops north of Isla Tiburon may be flooded (type 1) nor Tmr3 (type 2) of Santa Isabel inferred here to overlie caldera-collapse struc-
by the Gulf of California or possibly were re- Wash may be correlative to Tmr3 units tures (Fig. 6). This same region encloses an
moved by erosion. mapped by Lewis and Stock (1998a). Reex- area of extensive rhyolite lava flows and late
amination of individual sample localities re- Miocene and younger andesite volcanoes that
The Tuffs of Mesa Cuadrada veals that the range-wide averaging technique probably filled in the collapsed area (Oskin
used by Nagy (2000) to test these correlations and Stock, 2003b).
The tuffs of Mesa Cuadrada (Figs. 5 and 6) included more samples of Tmr4 from a zone
comprise two cooling units, Tmr3 and Tmr4, of greater clockwise vertical-axis rotation in Tuffs of Dead Battery Canyon
with intercalated Tmr3t nonwelded pyroclastic the southern Sierra San Fermn (Lewis, 1994).
deposits. Tmr3 and Tmr4 form a distinctive This approach biased the average declination The tuffs of Dead Battery Canyon (Figs. 7
pair of flow deposits with contrasting litholo- for Tmr4 relative to the average declination of and 8) in Baja California form two nearly
gy and paleomagnetic remanence (Table 1). Tmr3 units from the Sierra San Fermn; the identical cooling units, Tmr5 and Tmr6 (Table
Geochemical analyses (Table 2) show that, result was that the Tmr3 units of Lewis and 1). Pumice and phenocryst content of outcrops
similar to the tuff of San Felipe, cooling unit Stock (1998a) appeared paleomagnetically on Isla Tiburon best supports correlation to
Tmr3 is also higher in Nb, Ce, Rb, and La distinct from the Tmr3 of Santa Isabel Wash. cooling unit Tmr5. Cooling units Tmr5 and
than any other silicic rocks from the Puerte- By using individual sample localities from the Tmr6 in Baja California both preserve normal-

Geological Society of America Bulletin, October 2003 1177


Figure 3. Field photographs of the tuff of San Felipe. (A) Basal vitrophyre and densely
welded tuff, Sierra Menor, Isla Tiburon. The tuff of San Felipe is 85 m thick at this
locality. (B) Close-up of high-grade, densely welded tuff with light-colored pumice and
dark rhyolite lithic fragments, Sierra San Felipe, Baja California. Lens cap is 5.5 cm in
diameter.
polarity, moderate-inclination paleomagnetic
remanence directions; those of Tmr6 are
steeper and clockwise of unit Tmr5 (Lewis
and Stock, 1998a). Paleomagnetic remanence
of the one cooling unit present on Isla Tiburon
is consistent with the overall paleomagnetic
directions of Tmr5 and Tmr6 from Baja Cal-
ifornia, but it is not distinctive of either Tmr5
or Tmr6 given the uncertainties present in the
data (Table DR1 [see footnote 1]).
The tuffs of Dead Battery Canyon may be
an uppermost cooling unit of the tuffs of Mesa
Cuadrada. This possibility was suggested by
Lewis (1994), who observed that the tuffs of
Dead Battery Canyon are lithologically simi-
lar to and conformably overlie Tmr4 in the
Sierra San Fermn. Younger units, including a
possible correlative to the tuffs of Arroyo El
Canelo (Tmr8, Table 1) overlie the tuffs of
Dead Battery Canyon with a slight angular
unconformity (Lewis, 1994). This same rela-
tionship appears on Isla Tiburon at Punta Rei-
1178 Geological Society of America Bulletin, October 2003
OSKIN and STOCK
na where the tuffs of Dead Battery Canyon
conformably overlie the tuffs of Mesa Cuad-
rada but are overlain with slight angular un-
conformity by the El Oculto member of the
tuffs of Arroyo El Canelo (Fig. 7). Although
the tuffs of Dead Battery Canyon have not
been dated, these relationships suggest an age
close to the ca. 6.3 Ma Tmr3 cooling unit of
the tuffs of Mesa Cuadrada.
The tuffs of Dead Battery Canyon cover a
restricted area on both margins of the northern
Gulf of California (Fig. 8). The maximum
thickness of these tuffs occurs where two
cooling units (Tmr5 and Tmr6) are present in
Baja California. Thick deposits of the Tmr5
cooling unit are present in the Punta Reina
area of Isla Tiburon. The tuffs of Dead Battery
Canyon have the lowest volume of any unit
that has been found to correlate across the
northern Gulf of California. Known outcrops
of the tuffs of Dead Battery Canyon cover 160
km2 with ;4 km3 of pyroclastic deposits (Ta-
ble 3). The northern limit of the tuffs of Dead
Battery Canyon is well exposed in the central
Sierra San Fermn of Baja California and the
northern Sierra Menor of Isla Tiburon. These
ignimbrites thin southward to zero thickness
in the Sierra Menor and are buried by younger
rhyolite lava flows in the southernmost Sierra
San Fermn (Fig. 8). The tuffs of Dead Battery
Canyon do not appear to extend south of Ar-
royo Matom in Baja California.
The Tuffs of Arroyo El Canelo
The tuffs of Arroyo El Canelo are a com-
pletely zoned ignimbrite with at least four dis-
tinct cooling units of a complexly zoned ig-
nimbrite (Figs. 7 and 9; Table 1). The
stratigraphic definition of these tuffs is based
upon the exposures in the Santa Isabel Wash
region (Nagy et al., 1999). The units desig-
nated as members of the tuffs of Arroyo El
Canelo in this study were given separate
names by Nagy et al. (1999). These litholog-
ically distinctive members are the tuff of Ar-
royo El Oculto, the tuff of Arroyo El Canelo
(with one or more cooling breaks), the Big-
horn Sheep Tuff, and the Flagpole Tuff (Table
1). Correlation of the tuffs of Arroyo El Ca-
nelo has not been confirmed by geochemical
or paleomagnetic analyses. However, the com-
plex zonation of the tuff of Arroyo El Canelo
is consistent between the Sierra Santa Isabel
of Baja California and western Isla Tiburon.
Plagioclase phenocrysts from the El Canelo
member from Santa Isabel Wash were isoto-
pically dated as 6.1 6 0.5 Ma by Nagy et al.
(1999).
The four cooling units of the tuffs of Ar-
royo El Canelo cover .700 km2 with ;45
km3 of pyroclastic deposits (Fig. 9 and Table
3). Outcrops of the tuffs of Arroyo El Canelo
occur south of the tuffs of Dead Battery Can-
yon and partly overlap, but extend southeast
of, the tuffs of Mesa Cuadrada. This pattern
is seen on both margins of the Gulf of Cali-
fornia (Figs. 6, 8, and 9). In Baja California,
the tuffs of Arroyo El Canelo crop out pri-
marily south of Arroyo Matom, except for
isolated exposures in the southern Sierra San
Fermn. On Isla Tiburon, the tuffs of Arroyo
El Canelo crop out south of Punta Reina and
within the central and southern Sierra Menor.
On both margins, the northern outcrop limit
of the tuffs of Arroyo El Canelo lies south of
the northern outcrop limit of the tuffs of Dead
Battery Canyon. Also, the tuffs of Arroyo El
Canelo crop out farther southeast than all oth-
er correlated ignimbrites.
Thicker, high-grade welded deposits of the
tuffs of Arroyo El Canelo (e.g., tuff that is
 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

Figure 4. Distribution of the tuff of San Felipe in (left) Baja California and (right) coastal Sonora and Isla Tiburon. Tuff outcrops
shown in black; stratigraphic thickness measurements in meters shown as numbers. Isopachs of tuff thickness shown as thin black lines
surrounding gray patches. Darker shades of gray indicate thicker deposits: 30 m contour intervals show thicknesses up to 90 m;
thicknesses above 90 m not contoured. Dashed and dotted lines reflect uncertainty in isopach positions. Outcrops of the tuff of San
Felipe fill west-trending paleocanyons on Isla Tiburon and the southern quarter of the outcrop area in Baja California. Thicker outcrops
of the tuff of San Felipe on coastal Sonora are related to a probable vent located east of Punta Chueca. The boundary of isopachs in
coastal Sonora was arbitrarily placed east of all mapped outcrops of the tuff of San Felipe on coastal Sonora. Additional deposits of
the tuff of San Felipe probably exist east of the mapped area.

densely welded throughout the section with
post-emplacement rheomorphic flow linea-
tions) form a more restricted distribution cen-
tered on outcrops of intra-caldera deposits in
the eastern Arroyo Matom area (Fig. 9). Pa-
leotopography formed during eruption of the
tuffs of Mesa Cuadrada (caldera wall(?) in
Figs. 6 and 9) acted as a buttress against
which the tuffs of Arroyo El Canelo were de-
posited. On Isla Tiburon, thick deposits of the
tuffs of Arroyo El Canelo crop out in two ar-
eas adjacent to the western coastline. Thick
high-grade deposits occur in the southwest Si-
erra Menor, south of a paleotopographic bar-
rier that was the southern limit of the tuffs of
Mesa Cuadrada (Fig. 6). The tuffs of Arroyo
El Canelo may have ponded here on the south
side of this barrier. Outcrops of the thick,
moderately to densely welded tuff of Arroyo
El Canelo occur adjacent to the coastline of
Isla Tiburon from the central Sierra Menor
northward to Punta Reina, where outcrops of
this ignimbrite abruptly terminate (Fig. 9).
These thick deposits are separated by normal
faults and buttress unconformities from thin-
ner sections to the east.
DISCUSSION
Pyroclastic flow deposits offset across the
northern Gulf of California permit restoration
of PacificNorth America displacement (Figs.
10, 11, and 12). This analysis builds on the
initial study of Oskin et al. (2001), which pro-
posed 255 6 10 km of opening of the Upper
Delfn basin in the northern Gulf of California
since the 6.3 Ma time of emplacement of
Tmr3 (e.g., Fig. 6). Alternative restorations of
245 and 265 km illustrate the robustness of
this reconstruction (Figs. 10A and 10C). Dis-
tributed dextral displacement of ignimbrite
distributions is shown to support 41 6 13 km
of additional plate-boundary motion between
Baja California and mainland Sonora (Fig.
11). By comparison of these results to plate-
motion circuit data, Oskin et al. (2001)
showed that the PacificNorth America plate
boundary became localized in the Gulf of Cal-
ifornia during latest Miocene time. A more
complete compilation of these data presented
here, which includes distributed deformation

Geological Society of America Bulletin, October 2003 1179

 OSKIN and STOCK

TABLE 3. VOLUME ESTIMATES OF CORRELATIVE ASH-FLOW TUFFS of the continental margin, further restricts the
Unit, location Area enclosed by contour Maximum Volume time of localization of the plate boundary. The
(km2) thickness (km3) record of plate-boundary motion from the
(m)
0m 30 m 60 m 90 m Gulf of California is shown to place signifi-
Tuff of San Felipe cant boundary conditions on models of the
Baja California 3360 1390 540 220 180 118 San Andreas fault system in southern Califor-
Isla Tiburon 440 300 60 85 17 nia. Close restoration of conjugate rift margins
Coastal Sonora 330 280 220 170 135 25
also restricts tectonic models for crustal evo-
Total: 160 lution of the Upper Delfn ocean basin.
Tuffs of Mesa Cuadrada
Baja California 1790 1270 850 330 110 98
Isla Tiburon 350 250 160 90 190 22 Map-View Restoration of Conjugate
Total: 120 Rifted Margins
Tuffs of Dead Battery Canyon
Baja California 110 30 20 74 3
Isla Tiburon 50 10 38 1 Map-view restoration of conjugate rifted
Total: 4 margins of the northern Gulf of California re-
Tuffs of Arroyo El Canelo quires 255 6 10 km of opening of the Upper
Baja California 590 290 230 180 325 41 Delfn basin since eruption of the tuffs of
Isla Tiburon 150 50 10 60 4
Mesa Cuadrada at 6.3 6 0.2 Ma (Oskin et al.,
Total: 45 2001) (Fig. 10). The best-fit restoration is sat-

Area measurements do not account for extension of the outcrop area after deposition and may overestimate isfied solely by motion parallel to the Tiburon
the volume of tuffs by up to a factor of two. Fracture Zone and Ballenas transform fault in

The 0 m contour was multiplied at 1/2 contour interval (15 m). Each complete contour above 0 m was multiplied
at 1 contour interval (30 m). Final contour interval was multiplied at 1/2 contour interval (15 m) 1 1/3 remaining the Gulf of California (3108 to 3128, Fig. 2;
thickness. Fenby and Gastil, 1991). The actual path of

The area and the volume estimate presented here account only for outcrops west of the coastal zone. Thick
outcrops of the Tuff of San Felipe in coastal Sonora indicate that there is likely to be additional tuff farther east.
opening between of the Upper Delfn basin
may have involved other components of mo-
tion. However, these other components must
be much less significant than motion parallel
to transform faults formed and lengthened by
opening of the Upper Delfn basin (Fig. 2). To

Figure 5. Field photographs of the tuffs of Mesa Cuadrada. (A) Close-up of the basal vitrophyre of the Tmr4 cooling unit, overlain by
densely welded tuff affected by spherulitic alteration, Sierra Menor, Isla Tiburon. (B) Tmr3 and Tmr4 cooling units of the tuffs of Mesa
Cuadrada unconformably overlying an eroded remnant of the tuff of San Felipe, Punta Reina, Isla Tiburon. Note that multiple cooling
horizons have developed in thick Tmr3 at this locality. Sea cliff is ;200 m high. (C) Close-up of Tmr3 outcrop, showing crystal- and
lithic-rich composition, Sierra San Fermn, Baja California. Lens cap is 5.5 cm in diameter.

1180 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

Figure 6. Distribution of the tuffs of Mesa Cuadrada in (left) Baja California and (right) Isla Tiburon. See Figure 4 for explanation of
isopach map symbols. Enclosed areas of nondeposition of the tuffs of Mesa Cuadrada correspond to paleotopographic highs. Adjacent
paleocanyons and basins are filled with as much as 190 m of the tuffs of Mesa Cuadrada. Tuff thickness changes abruptly near proposed,
partially buried caldera-wall features in Baja California. The eastern edge of tuff distribution was probably controlled by fault-bounded
relief on the west side of the Sierra Kunkaak. Isopachs of tuff thickness are inferred to terminate against this boundary. Rhyolite sample
1012 of Gastil et al. (1979) probably came from an outcrop of Tmr3 on Isla Tiburon, as shown by circled star.

account for rotation of 4.08 about the modified
NUVEL-1A PacificNorth America Euler
pole (lat 50.58N, long 75.88W, DeMets and
Dixon, 1999), Sonora and Isla Tiburon are ro-
tated 2.38 counterclockwise in map view rel-
ative to true north. The restoration path de-
fined by the Tiburon Fracture Zone and
Ballenas transform fault is slightly counter-
clockwise of a 3148 vector of PacificNorth
America plate motion at Puertecitos calculated
with this same Euler pole. Distributed defor-
mation of the Baja California margin (de-
scribed in Lewis and Stock, 1998a) may ac-
count for this difference, as discussed later in
the following section.
The amount and uncertainty of the 255 km
restoration (Fig. 10) were determined by the
limits of a reasonable fit between tie lines
from the tuffs of the northern Puertecitos vol-
canic province. The best fit of these tie lines
is depicted on the map-view restoration of 255
km of displacement (Fig. 10B). Fits with 245
km and 265 km of displacement parallel to the
transform faults illustrate the limits of the res-
toration (Figs. 10A and 10C). The partitioning
of dextral displacement between the Ballenas
transform fault and the Tiburon Fracture Zone
(Fig. 2) depends upon the restoration of Isla
Angel de la Guarda to an uncertain position
south of Isla Tiburon (Lonsdale, 1989; Stock,
2000). This uncertainty does not affect the
summed displacement measured here.
The distribution of the ca. 6.3 to 6.1 Ma
tuffs constrains the amount and uncertainty of
dextral displacement across the northern Gulf
of California since late Miocene time (Fig.
10). In particular, several stratigraphic transi-
tions in the Arroyo Matom and southern Si-
erra San Fermn region of Baja California cor-
relate to the Punta Reina region of Isla
Tiburon (Figs. 6, 8, and 9). Here, the tuffs of
Dead Battery Canyon and the tuffs of Arroyo
El Canelo abruptly pinch out above the tuffs
of Mesa Cuadrada. South of this tie line, thick,
high-grade welded, and intracaldera facies of
the tuffs of Mesa Cuadrada and the tuffs of
Arroyo El Canelo fill in the inferred eruptive
center at Arroyo Matom. The 255 km resto-
ration shows that thick deposits of these units
in the central and southern Sierra Menor orig-
inally were adjacent to the mouth of Arroyo
Matom. North of the southern Sierra San Fer-
mn and Punta Reina, only the tuffs of Mesa
Cuadrada are present. Both the thickness and
welding grade of cooling unit Tmr3 decrease
northward in the Sierra San Fermn of Baja
California and the Sierra Alta of Isla Tiburon.

Geological Society of America Bulletin, October 2003 1181

 OSKIN and STOCK

which is consistent with the radius of distri-

bution of the correlative ignimbrites.

Distributed Dextral Displacement

Map-view restoration of 255 6 10 km of

opening of the Upper Delfn basin between
Isla Tiburon and the coastline of northeastern
Baja California provides only a minimum es-
timate of the total dextral displacement across
the Gulf of California extensional province.
Outside of the Upper Delfn basin, distributed
deformation either west of the Baja California
coastline or east of Isla Tiburon may have
contributed to the total displacement across
the PacificNorth America plate boundary.
This section reviews existing stratigraphic and
paleomagnetic evidence to estimate the
amount of distributed dextral displacement on
both margin study areas (Fig. 11).
Two zones of discrete dextral displacement
are proposed for the northeastern Baja Cali-
fornia continental margin. The Valle de San
Felipe fault (Gastil et al., 1975; Dokka and
Merriam, 1982) (Fig. 11) may carry dextral
displacement from the Agua Blanca fault (Fig.
1) to a zone of distributed deformation in the
Puertecitos volcanic province. Detailed studies
Figure 7. (A) Field photograph and (B) interpretation of the tuffs of Mesa Cuadrada, tuffs of southern Valle Chico (Stock, 1993) corre-
of Dead Battery canyon, and the tuffs of Arroyo El Canelo near Punta Reina, Isla Tiburon, late volcanic strata of the Valle de San Felipe
showing angular unconformity between the tuffs of Arroyo El Canelo and slightly older fault, limiting significant dextral displace-
tuffs. Tmr3 and Tmr4tuffs of Mesa Cuadrada; Tmr5tuffs of Dead Battery Canyon; ment here. Similarly, the west- to northwest-
TmraoArroyo El Oculto member of the tuffs of Arroyo El Canelo; Tmrecmain (El trending Matom accommodation zone (Dok-
Canelo) members of the tuffs of Arroyo El Canelo with internal cooling break; Tmrfp ka and Merriam, 1982; Stock et al., 1991)
Flagpole member of the tuffs of Arroyo El Canelo; Tmalalluvium fill that unconform- (Fig. 11) appears to accommodate differential
ably overlies the tuff section. extension and rotation between the Puertecitos
volcanic province and the Sierra San Felipe
The distribution of the tuff of San Felipe distribution of older Tertiary conglomerate de- (Nagy, 2000) and does not carry significant
also matches in the favored map-view resto- posits (Figs. 2 and 10). Altogether, these re- dextral displacement.
ration (Figs. 6 and 10). The tuff of San Felipe sults suggest that no significant dextral motion Lewis and Stock (1998a) documented dis-
fills west-trending paleocanyons that crop out occurred between Isla Tiburon and Baja Cal- tributed dextral displacement within the San
along the entire western coastline of Isla Ti- ifornia prior to late Miocene time. Felipe rotated domain of the Baja California
buron. The Sierra San Fermn and the northern The detailed reconstruction permitted by margin (Fig. 11). Their paleomagnetic and
Puertecitos volcanic province contain similar matching the distributions of pyroclastic flow structural study proposed 23 6 9 km of dex-
paleocanyons filled by the tuff of San Felipe. deposits (Fig. 10) also limits the amount of tral displacement (along an azimuth of 3408)
The precision of the map-view restoration is upper continental crust on the continental mar- manifested by 308 6 158 of localized clock-
not enough to match individual paleocanyons; gins of the Upper Delfn basin that has sub- wise rotation since eruption of the ca. 6.3 Ma
however, high-grade welded deposits of the sided into the northern Gulf of California. The tuffs of Mesa Cuadrada (cooling units Tmr3
tuff of San Felipe in the northern Sierra Alta width of Miocene continental surface area not and Tmr4). The western boundary of the ro-
of Isla Tiburon and the northern Sierra San accounted for between Isla Tiburon and Baja tated domain is pinned to the Valle de San
Fermn of Baja California are restored into California is unlikely to exceed the 2030 km Felipe and/or San Pedro Martir faults. The
proximity. A once-continuous cover of the tuff radius of distribution of the tuffs of Mesa eastern boundary of the rotating domain is
of San Felipe in the Sierra San Felipe has no Cuadrada. This unit is the most widespread of proposed by Lewis and Stock (1998a) to lie
known correlative on coastal Sonora. Beneath the ca. 6.3 to 6.1 Ma tuffs (Fig. 6, Table 3). just offshore of northeastern Baja California.
the tuff of San Felipe, a stratigraphic pinch- Preservation of the tuffs of Dead Battery Can- This boundary is consistent with rocks on Isla
out of volcaniclastic strata of the early to mid- yon as a small-volume deposit on both rifted Tiburon, which are not significantly rotated,
dle Miocene arc against basement rocks may margins may require even less unaccounted and with faulting offshore of Baja California
also match in distribution on the map-view for crustal area. The gap between opposing (Persaud et al., 2003). When projected onto an
restoration (Oskin and Stock, 2003b). The pre- rifted margins shown in the preferred map- azimuth of 3148, parallel to plate motion,
ferred map-view restoration also matches the view restoration is ;20 to 25 km (Fig. 10), clockwise rotation of the Baja California mar-

1182 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

Figure 8. Distribution of the tuffs of Dead Battery Canyon in Baja California and Isla Tiburon. See Figure 4 for explanation of isopach
map symbols. Paleotopography and limited tuff volume probably restricted distribution of the tuffs of Dead Battery Canyon. Enclosed
area of zero thickness on Isla Tiburon indicates a paleotopographic high. A paleocanyon located northeast of this high and a basin
located southeast of this high each contains eastward-thinning outcrops of the tuffs of Dead Battery Canyon. Proposed caldera wall
corresponds to abrupt thinning of these ignimbrites from .30 m to ,5 m thickness in the southern Sierra San Fermn.

gin amounts to 21 6 9 km of displacement
(Lewis and Stock, 1998a). A vector sum of 23
6 9 km of dextral displacement at an azimuth
of 3408 with 255 6 10 km of opening of the
Upper Delfn basin at an azimuth of 3128
yields a combined vector of 276 6 13 km
(errors summed as root mean squares) at an
azimuth of 3148, parallel to the modified NU-
VEL-1A direction of DeMets and Dixon
(1999) for Puertecitos.
The La Cruz fault forms a zone of discrete
dextral displacement mapped through south-
ern Isla Tiburon (Fig. 11). Neither the timing
of motion nor the amount of offset across the
La Cruz fault are well constrained. The only
deposits to correlate across the fault zone are
post6 Ma marine rocks and pyroclastic flow
deposits (Oskin and Stock, 2003a). Deposits
from the middle Miocene volcanic arc on ei-
ther side of the La Cruz fault do not correlate,
which may indicate a large amount of dextral
and/or normal displacement. The La Cruz
fault is interpreted here as an early strand of
the Tiburon Fracture Zone that forms the
southern margin of the Upper Delfn segment separates the area of distributed shear on the
of the Gulf of California (Fig. 2). The block
south of the La Cruz fault was probably trans-
ferred across the Tiburon Fracture Zone dur-
ing the earliest stages of opening of the Upper
Delfn basin (Fig. 11). Slip along this fault
therefore should not be included as an addi-
tional component of the opening of the Upper
Delfn basin. exceed the ;300 km estimated for the Gulf
A northwest-trending zone of dextral shear
and rotation is inferred on the mainland Sono-
ran margin and northeastern Isla Tiburon (Fig.
11). Offset of high-grade welded tuff of San
Felipe between Punta Chueca and the northern
Sierra Kunkaak suggests 20 6 10 km
of dextral displacement and/or northwest-
directed extension across this zone. Basalt
flows exposed in both of these areas may also
correlate. Another zone of dextral displace-
ment, the northwest-striking Sacrificio fault,
southwest from the Seri block on the northeast
(Fig. 11). The total displacement across this
structure is difficult to estimate, because no
volcanic strata match across the fault. How-
ever, outcrops of a distinctive conglomerate in
the Sierra Seri preclude displacement on this
structure that would sum with other offsets to
of California by Gastil et al. (1973). Because
276 6 13 km and 20 6 10 km of displace-
ment (296 6 17 km total) is already accounted
for, strike-slip displacement on the Sacrificio
fault is unlikely to exceed a few tens of
kilometers.
Distributed dextral deformation may also be

Geological Society of America Bulletin, October 2003 1183

 OSKIN and STOCK

Figure 9. Distribution of the tuffs of Arroyo El Canelo in Baja California and Isla Tiburon. See Figure 4 for explanation of isopach
map symbols. Tuff thickness changes abruptly near proposed caldera wall features in Baja California. PHPaleotopographic high that
formed a buttress that blocked deposition of the tuffs of Mesa Cuadrada. The tuffs of Arroyo El Canelo overtopped this buttress and
were deposited further south than the tuffs of Mesa Cuadrada (compare to Fig. 6). The eastern edge of tuff on Isla Tiburon was probably
controlled by fault-bounded relief on the west side of the Sierra Kunkaah. Isopachs of tuff thickness are inferred to terminate against
this boundary.

inferred from paleomagnetic declination data
from Isla Tiburon and mainland coastal So-
nora (Fig. 11). Localities on Isla Tiburon show
minor, 58158 clockwise vertical-axis rotation
of outcrops of the tuff of San Felipe relative
to reference localities in Baja California.
Overall, however, sample localities on Isla
Tiburon show significantly less clockwise
vertical-axis rotation than sample localities in
Baja California. Therefore, no additional dex-
tral displacement has been added from rota-
tion of localities on western Isla Tiburon. Lo-
calities on coastal Sonora show 258408
clockwise vertical-axis rotation of outcrops of
the tuff of San Felipe. This rotation may have
been caused by right-lateral shear in the coast-
al zone between the Sacrificio fault and Isla
Tiburon. Dextral displacement accumulated
by rotation of these localities is tentatively at-
tributed to displacement of correlative out-
crops between Punta Chueca and the northern
Sierra Kunkaak by 20 6 10 km.
In summary, the total distributed dextral
displacement within the marginal continental
areas of the Gulf of California rift adds 41 6
13 km of offset to the 255 6 10 km measured
by map-view restoration of correlative ignim-
brite strata (Fig. 11). Offset of the tuff of San
Felipe from Punta Chueca to the northern Si-
erra Kunkaak indicates 20 6 10 km of dis-
placement. Vertical-axis rotations documented
by paleomagnetic studies add 21 6 9 km of
dextral displacement from the Baja California
margin (Lewis and Stock, 1998a). Additional
dextral displacement contributed by the Sac-
rificio fault is unknown at present. However,
outcrops of a distinctive conglomerate in the
Sierra Seri that correlate to outcrops in the
Santa Rosa basin in Baja California (Gastil et
al., 1973) preclude a large dextral offset across
the Sacrificio fault. On the basis of the strati-
graphic and paleomagnetic arguments pre-
sented here, the total amount of opening of the
Upper Delfn segment of the northern Gulf of
California is 296 6 17 km along an azimuth
of 3148, of which 255 6 10 km has been ac-
commodated within the ocean basin. This val-
ue is the same as that indicated by offsets es-
timated from Tertiary conglomerate (Gastil et
al., 1973) and from pre-Tertiary batholithic
and metamorphic rocks (Silver and Chappell,
1988; Gastil et al., 1991).
Timing of Dextral Displacement
Isotopic ages for the tuff of San Felipe, tuffs
of Mesa Cuadrada, and the tuffs of Arroyo El

1184 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

Figure 10. (B) Preferred and (A and C) alternative map-view restorations of conjugate rifted margins of the upper Delfn and Tiburon
basins in the northern Gulf of California. (B) Map view showing 255 km of dextral displacement restored along an azimuth of 3128
relative to Puertecitos, Baja California. This restoration best matches features near opposite shorelines of the northern Gulf of California
in an east-west alignment. (A and C) Alternative interpretations showing 265 km and 245 km of displacement restored along an azimuth
of 3128 relative to Puertecitos, Baja California. The 265 km restoration closely restores the edges of mapped outcrops but misaligns
distributions of the tuffs of Dead Battery Canyon and the tuffs of Arroyo El Canelo. The 245 km restoration reasonably aligns the
extent of correlative ignimbrites but requires a greater missing area of thick high-grade deposits between conjugate rift margins. Baja
California is held fixed in all frames. Coastal Sonora and Isla Tiburon are rotated 2.38 clockwise to account for finite rotation during
opening of the Gulf of California. Locations: CScoastal Sonora; PPuertecitos; PVPPuertecitos volcanic province; PRPunta
Reina; SASierra Alta; SFLSierra San Felipe; SFRSierra San Fermn; SMSierra Menor.

Canelo constrain the timing of dextral dis-
placement in the Gulf of California. Restora-
tion of northeastern Baja California to Isla Ti-
buron indicates that 255 6 10 km of opening
occurred in the Upper Delfn basin after em-
placement of the tuffs of Mesa Cuadrada at
6.3 6 0.2 Ma and the tuffs of Arroyo El Ca-
nelo at 6.1 6 0.5 Ma. Likewise, vertical-axis
rotations measured in the continental margin
of Baja California also postdate emplacement
of the tuffs of Mesa Cuadrada (e.g., Tmr3 and
Tmr4 of Lewis and Stock, 1998a). Together,
these displacements add up to 276 6 13 km
since ca. 6.3 Ma. Northwestward displacement
of 20 6 10 km between Isla Tiburon and the
Sacrificio fault accrued after emplacement of
the tuff of San Felipe at ca. 12.6 Ma.
Northeastern Baja California and Isla Ti-
buron do not show any evidence for signifi-
cant displacement on northwest-striking dex-
tral faults between 12.6 Ma and 6.3 Ma,
despite widespread evidence for extension
during this time period (Stock and Hodges,
1990; Lee et al., 1996; Lewis and Stock,
1998b). East- to northeast-directed extension
during this time accommodated a component
of PacificNorth America plate motion (cf.
Stock and Hodges, 1989). East-directed exten-
sion took place on Isla Tiburon (Oskin, 2002),
across northeastern Baja California (Stock and
Hodges, 1990; Lewis and Stock, 1998b), in-
cluding minor distributed dextral displacement
on the Matom accommodation zone (Nagy,
2000). However, the low magnitude of exten-
sion across Baja California and Isla Tiburon
(10% to 40% across a 100-km-wide zone)
does not approach the 160 6 80 km of east-
directed extension predicted by the plate-
circuit model of Stock and Hodges (1989).
Additional extension and right-lateral shear
within mainland Mexico probably makes up
the balance of strain necessary to complete the
plate circuit (Gans, 1997; Henry and Aranda-
Gomez, 1992).
In summary, timing information for dextral
displacement between Baja California and
mainland coastal Sonora supports substantial
dextral displacement of at least 276 6 13 km
from 6.36.1 Ma to the present. An additional
20 6 10 km of displacement occurred some-
time between 12.6 Ma and the present in a
region lacking dated rocks younger than 12.6
Ma. No evidence firmly supports substantial
dextral displacement within or significant
opening of the Upper Delfn basin segment of
the Gulf of California rift prior to 6.3 Ma, ex-
cept for a component accommodated by east-
directed continental extension from 12.6 Ma
to 6.3 Ma.
Dextral PacificNorth America Plate
Motion in the Northern Gulf of California
Restoration of conjugate rifted margins of
the Upper Delfn basin of the northern Gulf
of California defines the history of dextral
plate motion in the Gulf of California (Fig.
12). Finite rotations derived from a plate cir-
cuit (Atwater and Stock, 1998) define the full
PacificNorth America plate motion of ;760
km along an average azimuth of 3048 from
15.1 Ma (chron 5b, time scale of Cande and
Kent (1995)) to the present. Prior to 12.5 Ma,
the Magdalena microplate separated the Pacif-
ic plate from the North America plate offshore
of southern Baja California (Mammerickx and
Klitgord, 1982). Subduction was dextral-
oblique during this time (Atwater, 1989), and
some component of dextral plate-boundary
motion may have been absorbed within the
North American continent. However, deposi-
tion of the tuff of San Felipe as a preexten-
sional marker indicates that plate-boundary
motion probably did not occur within the
northern Gulf of California area west of coast-
al Sonora prior to ca. 12.6 Ma. Since 12.5 Ma,

Geological Society of America Bulletin, October 2003 1185

 OSKIN and STOCK

Figure 11. Tectonic block diagram illustrating distributed deformation of the continental margins of the Upper and Lower Delfn basins.
See Figure 2 for present configuration. The Gulf of California is shown closed by 255 km between Isla Tiburon and Baja California
(see Fig. 10 for rationale). Closure of the Lower Delfn basin is by restoration of Isla Angel de La Guarda to an uncertain position
southeast of Puertecitos (Lonsdale, 1989; Stock, 2000). The La Cruz block is shown restored southeast of Isla Tiburon, adjacent to Isla
Angel de la Guarda. The original position of this block is poorly constrained. Paleomagnetic declination anomalies for the tuff of San
Felipe, relative to the Mesa Cuadrada (MC) paleomagnetic reference site, shown by circular plots. Table DR1 (see footnote 1) lists
paleomagnetic remanence values and declination anomalies. Clockwise rotation of 308 6 158measured from the tuff of San Felipe and
the tuffs of Mesa Cuadrada (Lewis and Stock, 1998a)defines the San Felipe rotated domain on the Baja California rifted margin and
accommodates to 23 6 9 km of dextral shear along a 3408 azimuth. Displacement of tuff of San Felipe outcrops between Isla Tiburon
and coastal Sonora resulted in 20 6 10 km of dextral displacement and rotation of tuff of San Felipe outcrops. Additional right-lateral
displacement on the Sacrificio fault is unknown. Paleomagnetic sample localities from Lewis and Stock (1998a), Stock et al. (1999), Nagy
(2000), and Oskin et al. (2001).

coast-parallel PacificNorth America plate
displacement has been divided into a compo-
nent of dextral displacement accommodated
across the Gulf of California, primarily after
6 Ma, and a second component of dextral dis-
placement accommodated outside of the Gulf
of California, primarily prior to 6 Ma (Fig.
12). The age and offset of correlative ignim-
brites across the northern Gulf of California
and the paleomagnetic record of seafloor
spreading in the Alarcon basin in the southern
Gulf of California (Fig. 1) define these com-
ponents. Continental extension in the Gulf ex-
tensional province accommodated a signifi-
cant component of the total plate motion
(Stock and Hodges, 1989) but contributed lit-
tle to opening of the Upper Delfn basin prior
to 6 Ma.
The paleomagnetic record of seafloor
spreading in the Alarcon basin defines the slip
history in the Gulf of California from 3.6 Ma
to present (Fig. 12). DeMets (1995) docu-
mented that the rate of seafloor spreading here
was 10%15% less than the full PacificNorth
America displacement rate prior to 1 Ma. This
lower rate is shown as a deviation of the rate
of motion across the Gulf of California from
the PacificNorth America plate-circuit rate
(Fig. 12). At 3.6 Ma, this discrepancy is sig-
nificant relative to the error for these recon-
structions. The rate of seafloor spreading in
the Gulf of California in the past 1 m.y. is
indistinguishable from the full PacificNorth

1186 Geological Society of America Bulletin, October 2003

 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA

terval at ca. 6 6 1 Ma. It is likely that the

large-volume silicic eruptions that produced
the tuffs of Mesa Cuadrada and the tuffs of
Arroyo El Canelo mark this localization event
(Oskin and Stock, 2003b). Additional dextral
displacement measured within coastal Sonora
and between coastal Sonora and Isla Tiburon
may be attributed to pre and/or post6 Ma
slip. However, the majority of PacificNorth
America plate-boundary motion from 12.5 Ma
to 6.3 Ma must have occurred outside of the
Gulf of California (Oskin et al., 2001) (Fig.
12).
The slip history presented here (Fig. 12)
provides a test of existing models for opening
of the Gulf of California. Lonsdale (1989)
proposed that 50 km of seafloor spreading be-
tween Baja California and North America oc-
curred at the Maria Magdalena Rise from 6 or
5 Ma until the onset of spreading at the Alar-
con Rise at 3.6 Ma. The 276 6 13 km of
offset measured across the northern Gulf of
California requires ;120 km more displace-
ment than is recorded in the Alarcon basin.
Seafloor spreading at the Maria Magdalena
Rise, with additional displacement on the con-
tinental margins of the mouth of the Gulf of
California during this time interval, would sat-
isfy the amount of opening measured across
the northern Gulf of California. Umhoefer et
al. (1994) proposed acceleration of the Baja
Figure 12. Dextral displacement in the Gulf of California relative to PacificNorth America CaliforniaNorth America displacement rate
plate motion. Lines, labeled with chron numbers, on graph indicate displacement with at 3.5 Ma. A detailed resolution of dextral dis-
time. Light gray fields indicate uncertainty. The preponderance of PacificNorth America placement in the Gulf of California prior to
plate motion since ca. 6 Ma occurred in the Gulf of California, as indicated by seafloor 3.5 Ma is not possible with the current geo-
magnetic lineations at the Alarcon Rise (Lonsdale, 1989; DeMets, 1995) (Fig. 1) and by logic data. However, the total amount of slip
offset of the tuffs of Mesa Cuadrada and the tuffs of Arroyo El Canelo. Restoration of recorded in the northern Gulf of California
the tuff of San Felipe indicates that PacificNorth America plate motion from 12.5 Ma to since ca. 6 Ma cannot accommodate signifi-
ca. 6 Ma occurred outside of the Gulf of California. This slip discrepancy may be absorbed cant acceleration of the displacement rate in
by dextral displacement west of the Baja California Peninsula (Spencer and Normark, the Upper Delfn basin at 1 Ma and 3.5 Ma
1979) and/or by additional displacement within the southern Basin and Range province (Oskin et al., 2001). In either case, the major-
(Gans, 1997). ity of PacificNorth America plate motion,
276 6 13 km, must have been accommodated
in the Gulf of California since at least 6.1 6
America rate within the uncertainty of recent seafloor spreading at the Alarcon basin back 0.5 Ma to 5.5 6 0.4 Ma.
reconstructions (DeMets and Dixon, 1999), al- to 6.6 6 0.6 Ma. This extrapolated age is in-
though neotectonic studies of western Baja distinguishable from isotopic ages of the tuffs Implications for Dextral PacificNorth
California Sur indicate minor activity (Fletch- of Mesa Cuadrada (6.3 6 0.2 Ma) and the America Plate Motion in Southern
er et al., 2000; Dixon et al., 2000). tuffs of Arroyo El Canelo (6.1 6 0.5 Ma). If California
Restoration of the correlative ignimbrites the full PacificNorth America rate of motion
across the northern Gulf of California (Fig. (DeMets and Dixon, 1999) were accommo- Models for the development of the Pacific
10) requires that the PacificNorth America dated in the Gulf of California, 276 6 13 km North America plate boundary in southern
plate boundary and most of its motion was of displacement could have accumulated in California support a variety of displacement
localized in the Gulf of California during or 5.5 6 0.4 m.y. Separation of outcrops of the histories for the Gulf of California (e.g., Dick-
soon after eruption of the tuffs of Mesa Cuad- ca. 12.6 Ma tuff of San Felipe from Isla Ti- inson, 1996). The slip history presented here
rada and tuffs of Arroyo El Canelo. Opening buron to Baja California does not require any for the northern Gulf of California (Fig. 12)
of the Upper Delfn basin since this time is at additional dextral offset here prior to 6.3 Ma. provides a new constraint with which to eval-
least 276 6 13 km (Fig. 12). To accumulate These results indicate that the PacificNorth uate these models independent of the tectonic
this displacement measured from the offset ig- America plate boundary became localized in complexity of southern California. Recent ef-
nimbrites requires extrapolation of the rate of the Gulf of California during a short time in- forts to summarize the amount of deformation

Geological Society of America Bulletin, October 2003 1187


in central California successfully reconcile
PacificNorth America motion measured from
continental-deformation and plate-circuit data
(Dickinson and Wernicke, 1997; Atwater and
Stock, 1998). Similarly, displacement and tim-
ing information from the Gulf of California
reconciles the majority of plate-boundary mo-
tion in northwestern Mexico since ca. 6 Ma
(Fig. 12). In southern California, these data
require that most of the dextral shear trans-
mitted through the Transverse Ranges to the
Salton Trough (Fig. 1) occurred after ca. 6 Ma.
Most models of the San Andreas fault sys-
tem in southern California are compatible with
offsets and timing measured in the northern
Gulf of California, despite a range of slip es-
timates for the southern San Andreas fault
from 150 to 180 km (Matti et al., 1992) to
240 km (Ehlig et al., 1975). When summed
with additional displacement for other struc-
tures away from the San Andreas fault trace,
both slip estimates overlap or exceed the 276
6 13 km of dextral displacement measured for
the northern Gulf of California since ca. 6 Ma.
The most significant of these additional dis-
placements are 12 6 2 km of slip from the
Elsinore fault (Hull and Nicholson, 1992), 45
75 km of shear by rotation of the eastern
Transverse Ranges (Richard, 1993; Dickinson,
1996), and possibly up to 22 km of slip on
the Agua Blanca fault (Allen et al., 1960). Ad-
dition of slip from the Agua Blanca fault to
the Gulf of California is problematic because
this fault appears to terminate before reaching
the Gulf extensional province (Allen et al.,
1960; Lee et al., 1996). A slip of 26 6 2 km
on the San Jacinto fault (Sharp, 1967) is in-
cluded with estimates for the San Andreas
fault.
If only 150180 km of slip has occurred on
the southern San Andreas fault (Matti et al.,
1992), the total dextral displacement measured
in the northern Gulf of California requires a
considerable amount of slip along the San Ga-
briel fault (Fig. 1). Adding 60 6 5 km of slip
measured for the western San Gabriel fault
(Crowell, 1974) yields a total of 289 6 18 to
300 6 21 km dextral displacement from
southern California that would have been
transferred into the Gulf of California (Dick-
inson, 1996). These values overlap the 296 6
17 km total-displacement estimate for the
northern Gulf of California. An acceptable
match is also obtained by adding only 22 6
1 km slip measured for the north branch of
the San Gabriel fault (Ehlig, 1981), yielding a
total of 269 6 13 to 280 6 16 km dextral
displacement transferred into the Gulf of Cal-
ifornia (Dickinson, 1996). These estimates
both support the conclusion that dextral dis-
1188 Geological Society of America Bulletin, October 2003
OSKIN and STOCK
placement in the Gulf of California from 12.5
Ma to 6.3 Ma probably does not exceed the
20 6 10 km estimate for displacement and
rotation of outcrops of the tuff of San Felipe
east of Isla Tiburon.
Alternatively, higher estimates of displace-
ment for the southern San Andreas fault (Eh-
lig et al., 1975; Powell, 1993) could require
up to 390 6 20 km of dextral displacement
within northwest Mexico (Dickinson, 1996).
To accommodate this slip would require ;100
km of dextral displacement east of the Gulf of
California, most of which would have had to
occur prior to ca. 6 Ma. Significant problems
remain in reconciliation of this higher slip val-
ue in southern California (Richard, 1993). Ex-
isting data from northwest Mexico cannot rule
out substantial dextral displacement east of the
Gulf of California during Miocene time. Ul-
timately, up to 350 km of dextral displacement
is required in northwest Mexico from 12.5 Ma
to 6.3 Ma (slip discrepancy in Fig. 12). Most
of this slip is usually assigned to dextral dis-
placement on the Tosco-Abreojos fault zone
west of Baja California and extension in the
Gulf extensional province (Spencer and Nor-
mark, 1979; Stock and Hodges, 1989) (Fig. 1).
However, observations from southern Califor-
nia and evidence for strike-slip faulting east
of Isla Tiburon (Gastil and Krummenacher,
1977) and in southeastern Sonora (Gans,
1997) indicate that additional field work is
necessary to address the distribution of dextral
shear during middle to late Miocene time.
Implications for Crustal Structure of the
Northern Gulf of California
Foundered upper continental crust probably
forms only a small component of the crust of
the Upper Delfn basin of the northern Gulf
of California. The proximity of restored con-
jugate rifted margins of the Upper Delfn ba-
sin segment of the rift indicates that the miss-
ing width of continental surface area is
unlikely to exceed 25 km (Figs. 10 and 11).
This result implies that ;1000% southeast-
directed extension has occurred within the Up-
per Delfn basin. Most likely, opening of the
Upper Delfn basin has been accomplished by
a combination of (1) large-magnitude exten-
sion of the foundered edge of the continental
margin, (2) strain partitioning through conju-
gate strike-slip faulting and/or northeast-
directed extension or to fill in the gap left by
opening of the basin, and (3) rupture of the
continent and formation of new transitional
oceanic crust. Because of the narrow width of
missing continental surface area, (1) signifi-
cant finite extension of the foundered conti-
nental margins of the Upper Delfn basin
would probably result in exhumation of lower-
crustal levels. Exhumation of middle to lower
continental crust as a metamorphic core com-
plex (cf. Gonzalez-Fernandez et al., 2000) and
exhumation of serpentinized continental-
mantle lithosphere (Nicholas, 1985) have been
proposed to have produced part of the basin
floor in the northern Gulf of California. (2)
Strain partitioning does not appear to be a suf-
ficient mechanism to compensate for opening
of the Upper Delfn basin because faulting of
the appropriate magnitude has not been de-
scribed from the ocean basin or its margins.
Strain partitioning as proposed by Nagy and
Stock (2000) requires substantially more sub-
sided continental margin than is indicated, al-
though variations on this model could involve
exhumed or transitional crust. Significant con-
jugate strike-slip faulting may also be expect-
ed to disrupt throughgoing transform faults in
the ocean basin, contrary to the existing in-
terpretation of the Tiburon Fracture Zone
(Fenby and Gastil, 1991) (Fig. 1). Present-day
crustal structure and volcanism in the northern
Gulf of California and Salton Trough indicate
recent crustal formation in the absence of con-
tinental crust and thus support possibility 3
that crustal rupture occurred at some time dur-
ing opening of the northern Gulf of California
(Herzig and Jacobs, 1994; Couch et al., 1991).
Our results from matching conjugate rifted
margins cannot determine if or when each of
these processes may have contributed to open-
ing of the Upper Delfn basin. However, our
results do require that any mechanism to open
the Upper Delfn basin must preserve most of
the surface area of the upper continental crust
on the basins rifted margins.
CONCLUSIONS
Four pyroclastic flow deposits, with a total
of eight cooling units, are correlated from
northeastern Baja California to Isla Tiburon
and coastal Sonora. These pyroclastic flow de-
posits are the ca. 12.6 Ma tuff of San Felipe,
the 6.3 6 0.2 Ma tuffs of Mesa Cuadrada
(units Tmr3 and Tmr4), the tuffs of Dead Bat-
tery Canyon (unit Tmr5), and the 6.1 6 0.5
Ma tuffs of Arroyo El Canelo (including the
El Oculto, El Canelo, Bighorn Sheep, and
Flagpole units). Restoration of the ca. 6.3 to
6.1 Ma tuffs indicates 255 6 10 km of dextral
displacement, at an azimuth of 3128, between
northeastern Baja California and western Isla
Tiburon. This restoration also satisfies the dis-
tribution of the tuff of San Felipe, supporting
the conclusion that almost all opening of the
Gulf of California took place within the past
 PACIFICNORTH AMERICA PLATE MOTION, NORTHERN GULF OF CALIFORNIA
6 m.y. Offset pyroclastic flows and paleomag- logical Society of America, Geology of North Amer-
ica, v. N, p. 2172.
netic declination anomalies support an addi- Atwater, T., and Stock, J.M., 1998, PacificNorth America
tional 41 6 13 km of distributed deformation plate tectonics of the Neogene southwestern United
States: An update: International Geology Review,
within northwest-trending zones on both mar- v. 40, p. 375402.
gins of the Upper Delfn basin, for a total of Bryant, B.A., 1986, Geology of the Sierra Santa Rosa ba-
296 6 17 km of opening measured across this sin, Baja California, Mexico [M.S. thesis]: San Diego,
San Diego State University, 75 p.
segment of the northern Gulf of California. At Burchfiel, B.C., Lipman, P.W., and Zoback, M.L., 1992,
least 276 6 13 km of this opening took place The Cordilleran orogen: Conterminous U.S.: Boulder,
after eruption of the ca. 6.3 to 6.1 Ma tuffs. Colorado, Geological Society of America, Geology of
North America, v. G-3, p. 724.
These results indicate that the PacificNorth Cande, S.C., and Kent, D.V., 1995, Revised calibration of
America plate boundary became localized in the geomagnetic polarity timescale for the Late Cre-
taceous and Cenozoic: Journal of Geophysical Re-
the Upper Delfn basin over a short time in- search, v. 100, p. 60936096.
terval at ca. 6 6 1 Ma and that prior to 6.3 Couch, R.W., Ness, G.E., Sanchez-Zamora, O., Calderon-
Ma, the majority of dextral plate-boundary Riveroll, G., Doguin, P., Plawman, T., Coperude, S.,
Huenn, B., and Gumma, W., 1991, Gravity anomalies
motion must have occurred outside of the and crustal structure of the Gulf and Peninsular prov-
northern Gulf of California. This slip history inces of the Californias, in Dauphin, J.P., and Simo-
also requires that most of the activity of the neit, B.R.T., eds., The Gulf and Peninsular provinces
of the Californias: American Association of Petroleum
southern San Andreas fault system in southern Geologists Memoir 47, p. 2545.
California occurred after 6 Ma. Restoration of Crowell, J.C., 1974, The San Gabriel fault and Ridge basin,
southern California, in Crowell, J.C., ed., San Andreas
conjugate rifted margins of the Upper Delfn fault in southern California: Sacramento, California
basin accounts for all but a 2025 km width Division of Mines and Geology Special Report 118,
of continental surface area. This result indi- p. 208221.
Crowell, J.C., 1981, An outline of the tectonic history of
cates that at least 1000% southeast-directed southeastern California, in Ernst, W.G., ed., The geo-
extension has occurred across this rift seg- tectonic development of California (Rubey Volume 1):
ment. Basically, the upper continental crust Englewood Cliffs, New Jersey, Prentice-Hall,
p. 583600.
probably ruptured near the present-day DeMets, C., 1995, A reappraisal of seafloor spreading lin-
coastlines. eations in the Gulf of California: Implications for
transfer of Baja California to the Pacific plate and es-
timates of PacificNorth America motion: Geophysi-
ACKNOWLEDGMENTS cal Research Letters, v. 22, p. 35453548.
DeMets, C., and Dixon, T.H., 1999, New kinematic models
Support was provided by National Science Foun- for PacificNorth America motion from 3 Ma to
dation grants EAR-9614674 and EAR-0001248 and present: I. Evidence for steady motion and biases in
a grant from the University of California MEXUS the NUVEL-1A model: Geophysical Research Letters,
program. We appreciate the support of Jaime v. 26, p. 19211924.
Roldan-Quintana and Carlos Gonzalez-Leon, of the Dickinson, W.R., 1996, Kinematics of transrotational tec-
Universidad Nacional Autonoma de Mexico. Per- tonism in the California transverse ranges and its con-
tribution to cumulative slip along the San Andreas
mission to enter Isla Tiburon was granted by the
transform fault system: Geological Society of America
Secretara de Medio Ambiente y Recursos Naturales Special Paper 305, 46 p.
and the Cumcaac (Seri) Indian Tribe. Prescott Col- Dickinson, W.R., and Wernicke, B.P., 1997, Reconciliation
lege Research Station, Baha Kino, generously pro- of San Andreas slip discrepancy by a combination of
vided logistical support during field studies. We are interior Basin and Range extension and transrotation
especially grateful to our Cumcaac guide, Ernesto near the coast: Geology, v. 25, p. 663665.
Molina. Scott Dobner, Matt Bachman, Robert Hous- Dixon, T., Farina, F., DeMets, C., Suarez-Vidal, F., Fletcher,
ton, Jason Wise, Naomi Marks, and Lesley Perg as- J., Marquez-Azua, B., Miller, M., Sancjez, O., and
sisted with field studies. Discussions with Gary Umhoefer, P., 2000, New kinematic models for
PacificNorth America motion from 3 Ma to present:
Axen, Arturo Martn-Barajas, Elizabeth Nagy, and
II. Evidence for a Baja California shear zone: Geo-
Claudia Lewis, and reviews by Paul Umhoefer and physical Research Letters, v. 27, p. 39613964.
John Fletcher contributed to the development of this Dokka, R.K., and Merriam, R.H., 1982, Late Cenozoic ex-
paper. California Institute of Technology Publication tension of northeastern Baja California, Mexico: Geo-
#8869. logical Society of America Bulletin, v. 93,
p. 371378.
Ehlig, P.L., 1981, Origin and tectonic history of the base-
REFERENCES CITED ment terrane of the San Gabriel Mountains, central
Transverse Ranges, in Ernst, W.G., ed., The geotec-
Abbott, P.L., and Smith, T.E., 1989, Sonora, Mexico, source tonic development of California (Rubey Volume 1):
for the Eocene Poway conglomerate of southern Cal- Englewood Cliffs, New Jersey, Prentice-Hall,
ifornia: Geology, v. 17, p. 329332. p. 253283.
Allen, C.R., Silver, L.T., and Stehlie, F.G., 1960, Agua Ehlig, P.L., Ehlert, K.W., and Crowe, B.M., 1975, Offset of
Blanca faultA major transverse structure of northern the upper Miocene Caliente and Mint Canyon For-
Baja California, Mexico: Geological Society of Amer- mations along the San Gabriel and San Andreas faults,
ica Bulletin, v. 71, p. 457482. in Crowell, J.C., ed., San Andreas fault in southern
Atwater, T., 1970, Implications of plate tectonics for the California: Sacramento, California Division of Mines
Cenozoic evolution of western North America: Geo- and Geology Special Report 118, p. 8392.
logical Society of America Bulletin, v. 81, Fenby, S.S., and Gastil, R.G., 1991, A seismo-tectonic map
p. 35133536. of the Gulf of California and surrounding areas, in
Atwater, T., 1989, Plate tectonic history of the northeast Dauphin, J.P., and Simoneit, B.R.T., eds., The Gulf
Pacific and western North America, in Winterer, E.L., and Peninsular provinces of the Californias: American
Hussong, D.M., and Decker, R.W., eds., The eastern Association of Petroleum Geologists Memoir 47,
Pacific Ocean and Hawaii: Boulder, Colorado, Geo- p. 7983.
Geological Society of America Bulletin, October 2003
Fletcher, J.M., Eakins, J., Sedlock, R.L., Mendoza-
Borrunda, R., Walter, R.C., Edwards, R.L., and Dixon,
T.H., 2000, Quaternary and Neogene slip history of
the Baja-Pacific plate margin: Baha Magdalena and
the southwestern borderland of Baja California: Eos
(Transactions, American Geophysical Union), v. 82,
p. 1232.
Gans, P.B., 1997, Large-magnitude OligoMiocene exten-
sion in southern Sonora: Implications for the tectonic
evolution of northwest Mexico: Tectonics, v. 16,
p. 388408.
Gastil, R.G., and Krummenacher, D., 1974, Reconnaissance
geologic map of coastal Sonora between Puerto Lobos
and Baha Kino: Geological Society of America Map
and Chart Series MC-16, scale 1:150,000.
Gastil, R.G., and Krummenacher, D., 1977, Reconnaissance
geology of coastal Sonora between Puerto Lobos and
Baha Kino: Geological Society of America Bulletin,
v. 88, p. 189198.
Gastil, R.G., Lemone, D.V., and Stewart, W.J., 1973, Perm-
ian fusulinids from near San Felipe, Baja California:
American Association of Petroleum Geologists Bul-
letin, v. 57, p. 746747.
Gastil, R.G., Phillips, R.P., and Allison, E.C., 1975, Recon-
naissance geology of the state of Baja California: Geo-
logical Society of America Memoir 140, 170 p.
Gastil, R.G., Krummenacher, D., and Minch, J.A., 1979,
The record of Cenozoic volcanism around the Gulf of
California: Geological Society of America Bulletin,
v. 90, p. 839857.
Gastil, R.G., Miller, R.H., Anderson, P., Crocker, J., Camp-
bell, M., Buch, P., Lothringer, C., Leier-Engeldardt, P.,
DeLattre, M., Hoobs, J., and Roldan-Quintana, J.,
1991, The relation between the Paleozoic strata on
opposite sides of the Gulf of California, in Perez-
Segura, E., and Jacques-Ayala, C., eds., Studies of
Sonoran geology: Geological Society of America Spe-
cial Paper 254, p. 716.
Gonzalez-Fernandez, A., Danobietia, J.J., Cordova, D.,
Delgado-Argote, L.A., Carbonell, R., Bartolome, R.,
and Michaud, F., 2000, Lithospheric structure of the
upper Gulf of California using deep seismic and grav-
ity data, in Magana, L.C., ed., MemoriasV Reunion
Internacional sobre la Geologa de la Pennsula de
Baja California: Loreto, Baja California Sur, Univer-
sidad Autonoma de Baja California, p. 6970.
Henry, C.D., and Aranda Gomez, J.J., 1992, The real south-
ern basin and range: Mid Cenozoic to Late Cenozoic
extension in Mexico: Geology, v. 20, p. 701704.
Herzig, C.T., and Jacobs, D.C., 1994, Cenozoic volcanism
and two-stage extension in the Salton Trough, south-
ern California and northern Baja California: Geology,
v. 22, p. 991994.
Hull, A.G., and Nicholson, C., 1992, Seismotectonics of the
northern Elsinore fault zone, southern California: Seis-
mological Society of America Bulletin, v. 82,
p. 800818.
Lee, J., Miller, M., Crippen, R., Hacker, B., and Ledesma-
Vazquez, J., 1996, Middle Miocene extension in the
Gulf extensional province, Baja California: Evidence
from the southern Sierra Juarez: Geological Society of
America Bulletin, v. 108, p. 505525.
Lewis, C.J., 1994, Constraints on extension in the Gulf ex-
tensional province from the Sierra San Fermn, north-
eastern Baja California, Mexico [Ph.D. thesis]: Bos-
ton, Massachusetts, Harvard University, 361 p.
Lewis, C.J., 1996, Stratigraphy and geochronology of Mio-
cene and Pliocene volcanic rocks in the Sierra San
Fermn and southern Sierra San Felipe, Baja Califor-
nia, Mexico: Geofsica Internacional, v. 35, p. 131.
Lewis, C.J., and Stock, J.M., 1998a, Paleomagnetic evi-
dence of localized vertical-axis rotation during Neo-
gene extension of the Sierra San Fermn, northeastern
Baja California, Mexico: Journal of Geophysical Re-
search, v. 103, p. 24552470.
Lewis, C.J., and Stock, J.M., 1998b, Late Miocene to recent
transtensional tectonics in the Sierra San Fermn,
northeastern Baja California, Mexico: Journal of
Structural Geology, v. 20, p. 10431063.
Lonsdale, P., 1989, Geology and tectonic history of the Gulf
of California, in Winterer, E.L., Hussong, D.M., and
Decker, R.W., eds., The eastern Pacific Ocean and Ha-
1189

waii: Boulder, Colorado, Geological Society of Amer-
ica, Geology of North America, v. N, p. 499521.
Mammerickx, J., and Klitgord, K.D., 1982, Northern East
Pacific Rise: Evolution from 25 m.y. B.P. to the pres-
ent: Journal of Geophysical Research, v. 87,
p. 67516759.
Martn-Barajas, A., and Stock, J.M., 1993, Estratigrafa y
petrologa de la secuencia volcanica de Puertecitos,
Noreste de Baja California, Transicion de un arco vol-
canico a rift, in Delgado-Argote, L.A., and Martn-
Barajas, A., eds., Contribuciones a la Tectonica del
Occidente de Mexico: Mexico City, Mexico, Mexican
Geophysical Union Monograph 1, p. 6689.
Martn-Barajas, A., Stock, J.M., Layer, P., Hausback, B.,
Renne, P., and Martnez-Lopez, M., 1995, Arc-rift
transition volcanism in the Puertecitos volcanic prov-
ince, northeastern Baja California, Mexico: Geological
Society of America Bulletin, v. 107, p. 407424.
Matti, J.C., Douglas, M.M., and Cox, B.F., 1992, The San
Andreas fault system in the vicinity of the central
Transverse Ranges province, southern California: U.S.
Geological Survey Open-File Report 92354.
Nagy, E.A., 1997, Extensional deformation and volcanism
within the northern Puertecitos volcanic province, Sierra
Santa Isabel, Baja California, Mexico [Ph.D. thesis]:
Pasadena, California Institute of Technology, 333 p.
Nagy, E.A., 2000, Extensional deformation and paleomag-
netism at the western margin of the Gulf extensional
province, Puertecitos volcanic province, northeastern
Baja California, Mexico: Geological Society of Amer-
ica Bulletin, v. 112, p. 857870.
Nagy, E.A., and Stock, J.M., 2000, Structural controls on
the continent-ocean transition in the northern Gulf of
California: Journal of Geophysical Research, v. 105,
p. 16,25116,269.
Nagy, E.A., Grove, M., and Stock, J.M., 1999, Age and
stratigraphic relationships of pre- and syn-rift volcanic
deposits in the northern Puertecitos volcanic province,
Baja California, Mexico: Journal of Volcanology and
Geothermal Research, v. 93, p. 130.
Nicholas, A., 1985, Novel type of crust produced during
continental rifting: Nature, v. 315, p. 112115.
Oskin, M., 2002, Tectonic evolution of the northern Gulf
of California, Mexico, deduced from conjugate rifted
margins of the Upper Delfn basin [Ph.D. thesis]: Pas-
adena, California Institute of Technology, 481 p.
Oskin, M., and Stock, J., 2003a, Marine incursion synchro-
nous with plate-boundary localization in the Gulf of
California: Geology, v. 31, p. 2326.
1190 Geological Society of America Bulletin, October 2003
OSKIN and STOCK
Oskin, M., and Stock, J., 2003b, Cenozoic volcanism and
tectonics of the continental margins of the Upper Del-
fn basin, northeastern Baja California and western
Sonora, in Johnson, S.E., Paterson, S., Fletcher, J.M.,
Girty, G., Kimbrough, D.L., and Martn-Barajas, A.,
eds., Tectonic evolution of northwestern Mexico and
the southwestern USA: Boulder, Colorado, Geological
Society of America Special Paper 374, (in press).
Oskin, M., Stock, J., and Martn-Barajas, A., 2001, Rapid
localization of PacificNorth America plate motion in
the Gulf of California: Geology, v. 29, p. 459462.
Persaud, P., Stock, J., Steckler, M., Martn-Barajas, A., Die-
bold, J., Gonzalez-Fernandez, A., and Mountain, G.,
2003, Active deformation and shallow structure of the
Wagner, Consag and Delfn basins, northern Gulf of
California, Mexico: Journal of Geophysical Research
(in press).
Powell, R.E., 1993, Balanced palinspastic reconstruction of
prelate Cenozoic paleogeology, southern California:
Geologic and kinematic constraints on evolution of the
San Andreas fault system, in Powell, R.E., Weldon,
R.J., II, and Matti, J.C., eds., The San Andreas fault
system: Displacement, palinspastic reconstruction, and
geologic evolution: Geological Society of America
Memoir 178, p. 1106.
Rhodes, J.M., 1988, Geochemistry of the 1984 Mauna Loa
eruption: Implications for magma storage and supply:
Journal of Geophysical Research, v. 93,
p. 44534466.
Richard, S.M., 1993, Palinspastic reconstruction of south-
eastern California and southwestern Arizona for the
middle Miocene: Tectonics, v. 12, p. 830854.
Sharp, R.V., 1967, San Jacinto fault zone in the Peninsular
Ranges of southern California: Geological Society of
America Bulletin, v. 78, p. 705729.
Silver, L.T., and Chappell, B.W., 1988, The Peninsular
Ranges batholith: An insight into the evolution of the
Cordilleran batholiths of southwestern North America:
Transactions of the Royal Society of Edinburgh: Earth
Sciences, v. 79, p. 105121.
Spencer, J.E., and Normark, W.R., 1979, Tosco-Abreojos
fault zone: A Neogene transform plate boundary with-
in the Pacific margin of southern Baja California: Ge-
ology, v. 7, p. 554557.
Stewart, J.H., 1998, Regional characteristics, tilt domains,
and extensional history of the later Cenozoic Basin
and Range province, western North America, in
Faulds, J.E., and Stewart, J.H., eds., Accommodation
zones and transfer zones: The regional segmentation
of the Basin and Range province: Geological Society
of America Special Paper 323, p. 4774.
Stock, J.M., 1989, Sequence and geochronology of Mio-
cene rocks adjacent to the main gulf escarpment:
Southern Valle Chico, Baja California Norte, Mexico:
Geofsica Internacional, v. 28, p. 851896.
Stock, J.M., 1993, Geologic map of southern Valle Chico
and adjacent regions, Baja California, Mexico: Geo-
logical Society of America Map and Chart Series,
scale 1:20,000, v. 76, p. 11.
Stock, J.M., 2000, Relation of the Puertecitos volcanic
province, Baja California, to development of the plate
boundary in the Gulf of California, in Delgado-
Granados, H., Aguirre-Diaz, G., and Stock, J.M., eds.,
Cenozoic tectonics and volcanism of Mexico: Geolog-
ical Society of America Special Paper 334,
p. 143155.
Stock, J.M., and Hodges, K.V., 1989, Pre-Pliocene exten-
sion around the Gulf of California and the transfer of
Baja California to the Pacific plate: Tectonics, v. 8,
p. 99115.
Stock, J.M., and Hodges, K.V., 1990, Miocene to Recent
structural development of an extensional accommo-
dation zone, northeastern Baja California, Mexico:
Journal of Structural Geology, v. 12, p. 315328.
Stock, J.M., and Molnar, P., 1988, Uncertainties and impli-
cations of the Late Cretaceous and Tertiary position
of North America relative to the Farallon, Kula, and
Pacific plates: Tectonics, v. 7, p. 13391384.
Stock, J.M., Martn-Barajas, A., Suarez-Vidal, F., and Mill-
er, M.M., 1991, Miocene to Holocene extensional tec-
tonics and volcanic stratigraphy of northeastern Baja
California, Mexico, in Walawender, M., and Hanan,
B., eds., Guidebook for the 1991 Annual Meeting of
the Geological Society of America, San Diego, Cali-
fornia: Boulder, Colorado, Geological Society of
America, p. 4467.
Stock, J.M., Lewis, C.J., and Nagy, E.A., 1999, The tuff of
San Felipe: An extensive middle Miocene pyroclastic
flow deposit in Baja California, Mexico: Journal of
Volcanology and Geothermal Research, v. 93,
p. 5374.
Umhoefer, P.J., Dorsey, R.J., and Renne, P., 1994, Tectonics
of the Pliocene Loreto basin, Baja California Sur,
Mexico, and evolution of the Gulf of California: Ge-
ology, v. 22, p. 649652.
MANUSCRIPT RECEIVED BY THE SOCIETY 18 MARCH 2002
REVISED MANUSCRIPT RECEIVED 26 DECEMBER 2002
MANUSCRIPT ACCEPTED 17 FEBRUARY 2003
Printed in the USA